The purpose of this Making Of is not only to show how we achieved the final shot for this animation, but also to point out the main problems we had to solve and why we choose one technique over another (Mov.01).
First off all we should start by pointing out that our main worries about making this shot were that we need to make a cool and stylish environment where we could place our 3D character ... but we couldn't spend too much time working on shaders or having hours and hours of rendering in HD resolution.
So we decided early in the process that we were going to go with a camera projection technique for the environment. This would give us a quick way of laying our painted environment over a base 3D model which our main character reacts to. Of course we knew this would pay off because we were going to have a lot of shots of this same environment over the episodes of the series, and doing projection would only leave us the character for the render stage which would help us to save a lot time.
The cons of going along with this technique was that we were a bit limited with the freedom of movement with the camera, so we had to really pay attention at the pre-production stage. We had to make a strong storyboard defining where the camera was going to be and how it was going to be moved around the scene, because in this particular case each different camera position needed a different background painting and projection. Therefore the projection technique and process is the core of this tutorial.
So once we were decided about the technique, we started with the concept design for this environment. We gathered all the reference material possible, from photographic books, illustrations, the internet, anything really.
The action takes place in a luthier's workshop, so we knew from the very first moment that the place needed need to be full of tools and instruments; that there should be things hanging all around. It should be chaotic and full of stuff, but also have a cozy feeling - as if it would be a great place to work in.
The illustrations were made and once we were satisfied with the result, we moved forward to the modeling stage (Fig.01 & Fig.02).
During the modeling stage we took the design to a full, fairly high level, model because even though we knew we were only going to use base or proxy models for projection at the end, we also knew that having the set all modeled and lit was going to give a lot of detailed information and a solid base for taking that full, still frame of the 3D environment to Photoshop for the painting process. If we hadn't modeled the full environment, we would have been faced with an incomplete, vague base mesh to paint over.
But it's all about making decisions based on how you can achieve the better result in the less amount of time possible, so even though we understood that making the whole model of the environment was going to take more time we were sure that it was going to help us a lot at the painting stage, and it did (Fig.03 - Fig.05)!